US7657028B2 - Method for classifying private information securely - Google Patents

Abstract

A method for securely classifying private data x of a first party Alice using a classifier H(x) of a second party Bob. The classifier is

$H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x\right)\right),$
where

$h_n\left(x\right)=\{\begin{array}{cc}{\alpha }_n& x^T{y}_n>{\Theta }_n\\ {\beta }_n& \mathrm{otherwise}\end{array},{\alpha }_n,{\beta }_n$
and Θ_n are scalar values and y_n is a vector storing parameters of the classifier. Bob generates a set of N random numbers, S₁, . . . , S_N, such that

$s=\sum _{n=1}^N{s}_n,$
for each n=1, . . . , N, the following substeps are performed: applying a secure dot product to x^Ty_n to obtain a_n for Alice and b_n for Bob; applying a secure millionaire protocol to determine whether a_n is larger than Θ_n−b_n, and returning a result of a_n+S_n, or β_n+S_n; accumulating, by Alice, the result in c_n. Then, apply the secure millionaire protocol to determine whether

$c=\sum _{n=1}^N{c}_n$
is larger than

$s=\sum _{n=1}^N{s}_n,$
and returning a positive sign if true, and a negative sign if false to classify the private data x.

Description

FIELD OF THE INVENTION

This invention relates generally to cooperative computing, and more particularly to performing cooperative computing in a secure manner using encryption techniques to classify private information.

BACKGROUND OF THE INVENTION

The Internet provides opportunities for cooperative computing. With cooperative computing, users and providers can exchange goods, services, and information. The Internet can also provide access to a classifier that can be used to classify data or signals.

Data classification in general is well known in the art. Of particular interest are binary classifiers. Such classifiers simply give a ‘yes’ or ‘no’ answer to indicate whether particular data belongs to a particular class or not.

Specifically, binary classification is the task of classifying objects into two groups on the basis of whether they have some predetermined property or not. Typical binary classification tasks include face recognition in images, medical testing of clinical data, and quality control of products. Generally, computer implemented classifiers automatically ‘learn’ a classification system. Well known methods suitable for learning binary classifiers include decision trees, Bayesian networks, support vector machines (SVM), and neural networks.

Private information classification (PIC) enables two parties, for example, Alice and Bob, to engage in a protocol that allows Alice to classify data using Bob's classifier without revealing anything to Bob (not even the classification result) and without learning anything about Bob's classifier, other than an answer to a classification request. PIC brings together the fields of machine learning and cooperative, multi-party computing, which is a sub-field of cryptography.

Secure multi-party computation is described by Yao, “How to generate and exchange secrets,” 27th FOCS, pp. 162-167, 1986. That gave a solution to a general two party problem. As a concrete example, consider the well known ‘millionaire problem’. Two parties want to find who has a larger number without revealing anything else about the numbers themselves.

Goldriech et al. extended the solution to n>2 parties, some of whom might be cheating, O. Goldreich, S. Micali and A. Wigderson, “How to play any mental game—a completeness theorem for protocols with honest majority,” 19th ACM Symposium on the Theory of Computing, pp. 218-229, 1987.

However, the original theoretical construct was too demanding to be of practical use. An introduction to Cryptography is given by B. Schneier, in Applied Cryptography, 1996, and a more advanced and theoretical treatment is given by O. Goldreich, in Foundations of Cryptography, 2004.

Since then, many secure protocols have been described for various applications. Relevant to the present invention are secure dot-products and oblivious polynomial evaluation, learning decision trees, and private information retrieval (PIR), Y. C. Chang and C. J. Lu, “Oblivious polynomial evaluation and oblivious neural learning,” AsiaCrypt: Advances in Cryptology. LNCS, Springer-Verlag, 2001; B. Chor, O. Goldreich, E. Kushilevitz and M. Sudan, Private Information Retrieval,” FOCS, 1995; Y. Lindell and B. Pinkas, “Privacy preserving data mining” Advances in Cryptology—Crypto2000, LNCS 1880, 2000; and M. Naor and B. Pinkas, “Oblivious Polynomial Evaluation,” Proc. of the 31st Symp. on Theory of Computer Science (STOC), pp. 245-254, May 1-4, 1999.

In a secure dot product, Alice and Bob respectively determine a dot-product of their private data vectors without revealing anything other than the result to each other. In some variants of the dot product protocol, Alice obtains the sum of the dot-product and some random number that is known only to Bob, while Bob learns nothing. This serves as a building block for more complex protocols.

In oblivious polynomial evaluation (OPE), Bob has a polynomial P(x) and Alice has a particular value x. Alice evaluates the polynomial at the value x without letting Bob know the value x. Bob does so, without revealing the polynomial.

The OPE has also been used for learning a decision tree where the training data are held by two parties. The parties want to jointly learn a decision tree without revealing their private data to each other. In the end, each party learns the decision tree that was trained using the combined data, but the private data of one party is not revealed to the other party.

PIC is an extension of private information retrieval (PIR). In PIR, Alice is interested in retrieving a data item from Bob's database without letting Bob know which element Alice selected. For example, Bob has a database of stock quotes and Alice would like to obtain the quote of a particular stock without letting Bob know which stock Alice selected. Bob is willing to let her do so. However, Bob wants to ensure that Alice can access one, and only one, stock quote.

A number of ways are known for reducing the communication and computation resources required by PIR, A. Beimel, Y. Ishai, E. Kushilevitz, and J.-F. Raymond, “Breaking the O(n1/(2k−1)) Barrier for Information—Theoretic Private Information Retrieval,” FOCS, 2002 and E. Kushilevitz and R. Ostrovsky, “Replication Is Not Needed: Single Database, Computationally-Private Information Retrieval,” FOCS 1997.

SUMMARY OF THE INVENTION

The invention provides a method for securely classifying private data x of a first party using a classifier H(x) of a second party. The classifier is

$H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x\right)\right),$
where

$h_n\left(x\right)=\{\begin{array}{cc}{\alpha }_n& x^T{y}_n>{\Theta }_n\\ {\beta }_n& \mathrm{otherwise}\end{array};{\alpha }_n,{\beta }_n,$
and Θ_n are scalar values; and y_n is a vector storing parameters of the classifier.

The second party generates a set of N random numbers: s₁, . . . , s_N, such that

$s=\sum _{n=1}^N{s}_n.$
For each n=1, . . . , N, the following substeps are performed: applying a secure dot product to x^Ty_n to obtain a_n for the first party and b_n for the second party; applying a secure millionaire protocol to determine whether a_n is larger than Θ_n−b_n, returning a result of a_n+s_n or β_n+s_n, and accumulating, by the first party, the result in c_n.

Then, the secure millionaire protocol is applied to determine whether

$c=\sum _{n=1}^N{c}_n$
is larger than

$s=\sum _{n=1}^N{s}_n$
which returns a positive sign if true and a negative sign if false to classify the private data x.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a flow diagram for classifying private data using a secure threshold classifier according to an embodiment of the invention;

FIG. 2 is a flow diagram for classifying private data using a secure polynomial classifier according to an embodiment of the invention;

FIG. 3 is a flow diagram for classifying private data using a secure Gaussian function classifier according to an embodiment of the invention;

FIG. 4 is a flow diagram for classifying private data using a secure sigmoid classifier according to an embodiment of the invention; and

FIG. 5 is a chart of the pseudo-code for classifying private data using a secure k-nn kernel classifier according to an embodiment of the invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

A first party, e.g., Alice, has a private data vector x, and a second party, e.g., Bob, has a trained ‘strong’ classifier, in the form of a function H(x). Typically, the strong classifier is a linear combination of weak classifiers h_n(x), see Y. Freund and R. E. Schapire, “A short introduction to boosting,” Journal of Japanese Society for Artificial Intelligence, 14(5), pp. 771-780, September 1999, incorporated herein by reference.

Alice and Bob would like to engage in a secure protocol that allows Bob to classify Alice's data vector x. In particular, both parties evaluate a classification function:

$\begin{array}{cc}H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x^T{y}_n\right)\right),& \left(1\right)\end{array}$
using cryptographic tools, where the vector x ∈ F^L and vectors {y_n}_n=1 ^N, where y_n ∈ F^L are L-dimensional vectors over a finite field F. The notation x^Ty_n is a dot product, as known in the art.

The vector y_n stores parameters of the classification function. Bob ‘learns’ the parameters y_n using conventional mechanisms for training a classifier for a particular classification task, such as data recognition and, in particular, face recognition in images or sound recognition in an audio signal. It is well known how to classify data and how to train classifiers for various tasks.

Alice has the private data x, and Bob has the parameters (N, h_n( ) and y_n). Eventually, Alice learns sign(H(x)) and nothing else, and Bob learns nothing about the Alice's private data vector x.

Specifically, we describe how to evaluate the following forms of the classification function h(x^Ty) in a secure manner.

A threshold classification function,

$h_n\left(x\right)=\{\begin{array}{cc}{\alpha }_n& x^T{y}_n>{\Theta }_n\\ {\beta }_n& \mathrm{otherwise},\end{array}$
where a and β and Θ are scalar values, e.g., 1 and 0, that are learned by Bob during training.

A polynomial classification function, h(x^Ty)=(x^Ty+c)^d, where c and d are scalar values known only to Bob.

A Gaussian classification function, h(x^Ty)=exp(γ∥x−y∥₂ ²), where γ is a scalar value known only to Bob.

A sigmoid classification function, h(x^Ty)=1/(1+exp(x^Ty)).

Note that the Gaussian function and sigmoid function can be approximated with a polynomial function. Here, we describe how to securely evaluate the Gaussian and sigmoid functions directly. In addition we describe how to construct a secure k-nn classifier.

First, we describe preliminary notations and some well known sub-protocols, and then we describe various secure classifiers H according to the embodiments of the invention.

Oblivious Polynomial Evaluation Protocol

We use the well known oblivious polynomial evaluation (OPE) protocol, see Y. C. Chang and C. J. Lu, “Oblivious polynomial evaluation and oblivious neural learning,” AsiaCrypt: Advances in Cryptology. LNCS, Springer-Verlag, 2001, incorporated herein by reference, and M. Naor and B. Pinkas, “Oblivious Polynomial Evaluation,” Proc. of the 31st Symp. on Theory of Computer Science (STOC), pp. 245-254, May 1-4, 1999, incorporated herein by reference.

Oblivious Transfer Protocol

An oblivious transfer (OT) enables Alice to select one element from a database of elements that Bob holds without revealing to Bob which element was selected, and without learning anything about the other elements in the database. The oblivious transfer protocol was described by Even et al. as a generalization of Rabin's ‘oblivious transfer’, S. Even, O. Goldreich and A. Lempel, “A Randomized Protocol for Signing Contracts,” Communications of the ACM 28, pp. 637-647, 1985, incorporated herein by reference, and M. O. Rabin, “How to exchange secrets by oblivious transfer,” Tech. Memo TR-81, Aiken Computation Laboratory, 1981, incorporated herein by reference.

For example, Bob has private elements M₀ and M₁, and Alice wants to select one of the elements without letting Bob know which element Alice selected. Bob is willing to let her do so provided that Alice does not learn anything about the other elements. The following protocol, based on RSA encryptions, can be used to solve the problem.

• • 1. Bob sends Alice two different public encryption keys K₀ and K₁.
  • 2. Alice generates a private key K, and encrypts her private key K with either of Bob's public encryption keys K₀ or K₁. For this example, Alice uses Bob's public key K₀. Alice sends Bob E(K, K₀), where E(K, K₀) denotes the encryption of Alice's private key by Bob's public key.
  • 3. Bob does not know which public key was used by Alice to encrypt her private key. Therefore, Bob decrypts E(K, K₀) with both of his private keys. Thus, he obtains both the real key K and an erroneous key K′.
  • 4. Bob sends Alice encryptions E(M₀, K) and E(M₁, K′). Alice decrypts the first of these messages with her private key K and obtains M₀.

Secure Dot Product Protocol

Before describing the details of the well known dot product protocol, we give an intuitive introduction. The goal is to partition a result of the dot product operation x^Ty into two shares a and b, where the share a is known only to Alice, and the share b is known only to Bob.

The dot product holds that x^Ty=a+b. We do this by breaking the product of every pair of elements x_i* y_i into two shares a_i and b_i, and then letting Alice and Bob sum the vectors a and b, respectively to obtain shares of the dot product. The protocol is described below.

Input: Alice has the vector x and Bob has the vector y.

Output: Alice and Bob have shares a and b, such that a+b=x^Ty.

• • 1. Bob generates a random vector b ∈ F^L.
  • 2. For each i=1, . . . , L, Alice and Bob perform the following sub-steps:
    • (a) Bob constructs a |F|-dimensional vector a, for all possible values of element x_i such that a_i=x_i*y_i−b_i
    • (b) Alice uses the OT₁ ^|F| with x_i as an index to select an element from the vector a. Alice stores the result in a_i.
  • 3. Alice and Bob sum their private vectors a and b, respectively, to obtain the shares a and b of the dot-product x^Ty.

Secure Millionaire Protocol

Alice and Bob would like to determine who has a larger number without revealing anything else about their numbers, see A. C. Yao, “How to generate and exchange secrets,” 27th FOCS, pp. 162-167, 1986, incorporated herein by reference. The millionaire protocol uses the OT protocol. The idea is to have Alice and Bob represent their numbers in a binary format, scan the binary numbers, one bit at a time, from the most significant bit to the least significant bit, and then obtain the result.

For each bit, Bob prepares a lookup table that is based on his current bit value and two possible bit values of Alice. Alice uses OT₁ ² to obtain some intermediate result, and both Alice and Bob continue to the next bit.

To solve the millionaire problem, we use a state variable s that can take one of three states:

1—Alice has a larger number,

2—Bob has a larger number, or

3—Undecided.

For each bit, Bob constructs a six-way lookup table that includes the three states of s and the two possible values of the next bit of Alice's number. An output is a next state after evaluating the current bit.

For example, if s=1 (Alice has a larger number), Bob's current bit is 1, and Alice's current bit is 0, then the output state is s=1 because Alice has the larger number, and Alice and Bob move to the next bit.

To prevent Alice from interpreting the state s, Bob can use a different permutation for each bit. For example, for the first bit, s=1 means that Alice has the larger number, but for the second bit, s=1 might mean that Bob has the larger number. The details are described below.

Input: Alice and Bob have non-negative numbers x and y, respectively.

Output: Alice and Bob learn if x>y without revealing the numbers x or y.

• • 1. Bob defines three states {A, B, U} that correspond respectively to: Alice has a larger number, Bob has a larger number, and undecided. For each bit, Bob encodes {A, B, U} using a different permutation of the numbers {1, 2, 3}.
  • 2. For the left most bit, Bob constructs a two-entry lookup table z⁽ⁿ⁾, such that

	yn = 0	yn = 1

xn = 0	U	B
xn = 1	A	U

• •  where y_n is the most significant bit of the number y. If y_n=0, then Bob should choose the left column, otherwise he should use the right column.
  • 3. Alice uses OT₁ ² with x_n as an index to obtain s⁽ⁿ⁾=z⁽ⁿ⁾(x_n).
  • 4. For each i=n−1, . . . , 1, Alice and Bob conduct the following sub-steps:
    • (a) Bob constructs a 6-entry lookup table z⁽ⁱ⁾ that is indexed by s⁽ⁱ⁾ and x_i, such that

	yi = 0	yi = 1

	s(i+1) = A xi = 0	A	A
	s(i+1) = B xi = 0	B	B
	s(i+1) = U xi = 0	U	B
	s(i+1) = A xi = 1	A	A
	s(i+1) = B xi = 1	B	B
	s(i+1) = U xi = 1	A	U

• • •  If y_i=0, then Bob should choose the left column, otherwise he should use the right column.
    • (b) Alice uses OT₁ ⁶ with s⁽ⁱ⁺¹⁾ and x_i as indices to obtain s⁽ⁱ⁾=z⁽ⁱ⁾(s⁽ⁱ⁺¹⁾, x_i).
  • 5. Bob sends Alice the meaning of the three states of s⁽¹⁾ of the least significant bit. Alice now knows which number is larger.

The above described basic, well known secure protocols are now used to describe the secure classification of private data.

Private Information Classification

FIG. 1 shows the steps of one embodiment of the invention.

Secure Threshold Classifier

Input: Alice has input data vector x, and Bob has a strong classifier

$H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x^T{y}_n\right)\right),$
where

$h_n\left(x\right)=\{\begin{array}{cc}{\alpha }_n& x^T{y}_n>{\Theta }_n\\ {\beta }_n& \mathrm{otherwise}\end{array}$
is a weak classifier, a_n, β_n and Θ_n are scalar values, e.g., 1 and 0, that are learned during training.

The vector y_n stores the parameters of Bob's classifier which Bob ‘learns’ using conventional training methods, such as decision trees, Bayesian networks, support vector machines (SVM), and neural networks.

Output: Alice obtains the result sign(H(x)) and nothing else, and Bob learns nothing about the data x.

• • 1. Alice provides data x 101, and Bob provides N threshold classifiers h_n(X), each with hyperplane y _n 102. Bob also generates 110 a set of N random numbers: s₁, . . . , s_N, such that

$s=\sum _{n=1}^N{s}_n.$

• • 2. For each n=1, . . . , N, Alice and Bob perform the following sub-steps:
    • (a) Alice and Bob obtain private shares a_n 121 and b _n 122, respectively, of the dot product x^Ty_n 120 using the secure dot-product protocol.
    • (b) Alice and Bob use the secure millionaire protocol to determine 130 which number is larger: a_n or Θ_n−b. Instead of returning a_n or β_n, the protocol returns as a result 131 either a_n+s_n, or β_n+s_n
    • (c) Alice stores the result in c _n 132.
  • 3. Alice and Bob use the secure millionaire protocol to determine 140 which of the numbers

$c=\sum _{n=1}^N{c}_n\mathrm{or}s=\sum _{n=1}^N{s}_n$

• •  is larger. The result is a sign 103. If Alice has a larger number, then x is classified positively, otherwise x is classified negatively.

Secure Polynomial Kernel Classifier

FIG. 2 shows the steps of one embodiment of the invention.

Input: Alice has data x, and Bob has the classifier

$H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x\right)\right),$
where h(x^Ty)=(x^Ty+c)^d, where c and d are parameters known only to Bob.

Output: Alice has the result sign(H(x)) and nothing else, and Bob learns nothing about the data x.

• • 1. Alice provides data x 101, and Bob provides N polynomial classifiers h_n(X), each with hyperplane y _n 102. Bob generates 110 a set of N random numbers: s₁, . . . , s_N, such that

$s=\sum _{n=1}^N{s}_n.$

• • 2. For each n=1, . . . , N, Alice and Bob perform the following sub-steps:
    • (a) Alice and Bob obtain private shares a_n 121 and b _n 122, respectively, of the dot product x^Ty_n 120, using the secure dot-product protocol.
    • (b) Bob constructs 230 a polynomial ƒ_n(a)=(a_n+b_n+c)^d+s_n=(x^Ty_n+c)^d+s_n.
    • (c) Alice and Bob use the OPE protocol to evaluate and determine 235 ƒ_n(a) and

$r=\sum _{n=1}^N{f}_n\left(a_n\right).$

• • 3. Alice and Bob use the secure millionaire protocol to determine 240 which number is larger: r or s. If Alice has a larger number, then x is classified positively, otherwise x is classified negatively.

Secure Gaussian Kernel Classifier

FIG. 3 shows the steps of one embodiment of the invention.

Input: Alice has input data x, and Bob has the classifier

$H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x\right)\right),$
where h_n(x^Ty)=exp(γ∥x−y∥₂ ²), a norm 2 or Euclidian distance, and γ is a scalar learned by Bob during training.

Output: Alice obtains the result sign(H(x)) and nothing else, and Bob learns nothing about the data x.

• • 1. Alice provides data x 101, Bob provides N polynomial classifiers h_n(x), each with hyperplane y _n 102.
  • 2. For each n=1, . . . , N, Alice and Bob conduct the following sub-steps:
    • (a) Bob selects a random number s_B and constructs 310 a function ƒ(z)=γz−s_B, where z is a scalar.
    • (b) Alice and Bob use OPE to obtain 320 a share s_A 321 for Alice, where s_A=ƒ(x^Tx), a secure dot product.
    • (c) Alice and Bob obtain 330 shares r_A 331 and r _B 332, respectively, of the dot product x^T(−2γy_n), using the secure dot-product protocol.
    • (d) Alice constructs t_A=exp(s_A)exp(r_A) 333.
    • (e) Bob constructs t_B=exp(s_B)exp(r_B)exp(γy_n ^Ty_n) 334.
    • (f) Alice and Bob obtain 340 private shares a_n 341 and b _n 342 of the dot product t_At_B=exp(s_A+r_A+s_B+r_B+γy_n ^Ty_n)=exp(γ∥x−y_n∥₂ ²), using the secure dot-product protocol.
  • 3. Alice determines an accumulation

$a=\sum _{n=1}^N{a}_{n}343,$

• •  and Bob determines an accumulation

$b=\sum _{n=1}^N{b}_{n}344.$

• • 4. Alice and Bob use the secure millionaire protocol to determine 350 the sign 103 that indicates which number is larger: a or b. If Alice has a larger number then x is classified positively, otherwise x is classified negatively.

Secure Sigmoid Kernel Classifier

Input: Alice has data x 101, and Bob has the classifier

$H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x\right)\right),$
where h(x^Ty)=1/(1+exp(x^Ty)).

Output: Alice has the result sign(H(x)) and nothing else, and Bob learns nothing about the data x.

• • 1. For each n=1, . . . , N, Alice and Bob perform the sub-steps:
    • (a) Alice and Bob obtain 120 shares a_n 121 and b _n 122, respectively, of the dot product x^Ty_n using the secure dot-product protocol, where y _n 102 is a vector supplied by Bob, as before.
    • (b) Alice and Bob determine 430 private shares s_A 431 and s _B 432 such that s_A+s_B=exp(a)exp(b) using the secure dot-product protocol.
    • (c) Bob selects r_B at random, and defines 440 ƒ(s_A)=r_B+r_B(s_A+s_B)=r_B(1+exp(x^Ty_n)).
    • (d) Alice and Bob use OPE to determine 450 r_A=ƒ(s_A).
    • (e) Alice and Bob determine 460 shares t_{A, n} and t_{B, n} of the dot-product

$\frac{1}{r_A}{r}_B=\frac{1}{r_B\left(1+\exp \left(x^T{y}_n\right)\right)}$

• • •  using the dot-product protocol.
  • 2. Alice determines

$t_A=\sum _{n=1}^N{t}_{A,n}471.$

• • 3. Bob determines

$t_B=\sum _{n=1}^N{t}_{B,n}472.$

• • 4. Alice and Bob use the secure millionaire protocol to determine 480 the sign 103 that indicates which number is larger: t_A or −t_B. If Alice has a larger number, then x is classified positively, otherwise x is classified negatively.

Secure k-nn Classifier

This protocol is a secure realization of a k-nn classifier. Alice has data x. Bob has points {y_i}_i=1 ^N with associated labels c(y_i). Given the data x, the object is to find the label of the majority of the k vectors y_i that are nearest to the data x. Here, we focus on a binary classification. Alice and Bob have additive shares of a radial distance r, such that within the distance r, there are exactly k points y_{i 1} , . . . ,y_{i k} .

Alice and Bob can use a radius protocol, described below, to privately count the number of values and use their labels to privately determine correct label for the data x. However, the k-nn classifier is defined in terms of k and not the radius r, so we provide a protocol that can determine the radius r, given k.

To do this, we define r_k to be the radius of the k^th furthest point from the data x, and describe how to find r_k, given k. After the radius r_k is obtained, we can determine the label of x. We assume that Alice and Bob can determine private shares of a squared Euclidean distance d(x, y_i) between the point x and any of the points y_i using the secure dot-product protocol discussed earlier.

The basic steps of the methods are shown in FIG. 5 in terms of pseudo code.

Input: Alice has input data x and a private share r_A of a radius r, and Bob has points y_i and a private share, r_B of the radius r.

Output: Alice learns a classification label of x, and nothing else, and Bob learns nothing.

• • 1. For each point y_i, Alice and Bob perform the following sub steps:
    • (a) Alice and Bob obtain private shares a and b of the dot product (−2x)^Ty.
    • (b) Bob defines s_B=b+y^Ty and a function ƒ(z)=z+s_B.
    • (c) Alice and Bob use OPE to evaluate s_A=ƒ(x^Tx+a), where s_A is a private share of Alice.
    • (d) Bob selects u_i at random.
    • (e) Alice and Bob compare s_A−r_A to r_B−s_B using the millionaire protocol. This is equivalent to comparing the squared Euclidean distance d(x, y_i) to the radius r. If s_A−r_A<r_B−s_B AND c(y_i)=1, then Bob returns u_i+1, otherwise Bob returns u_i. Alice stores the result in v_i.
  • 2. Alice and Bob determine

$v=\sum _{i=1}^N{v}_i$

• •  and

$u=\sum _{i=1}^N{u}_i,$

• •  respectively.
  • 3. Alice and Bob compare v and u+k/2 using the millionaire protocol. If v>u+k/2, then x is labeled 1, otherwise the label is 0.

Radius Protocol

Input: Alice has input data x, and Bob has a list of points y_i and a value k.

Output: Alice has a share r_A of the radius r of k nearest neighbors of x, and Bob has a share r_B, such that the radius r=r_A+r_B. Bob learns nothing about the data x.

• • 1. Alice and Bob select randomly r_A and r_B, respectively.
  • 2. Do, until termination:
    • (a) For each point y_i, Alice and Bob perform the following sub-steps:
      • i. Alice and Bob obtain shares a and b of the dot product (−2x)^Ty_i.
      • ii. Bob defines s_B=b+y_i ^Ty_i and a function ƒ(z)=z+s_B.
      • iii. Alice and Bob use OPE to evaluate s_A=ƒ(x^Tx+a), where s_A is a private share of Alice.
      • iv. Bob selects u_i at random.
      • v. Alice and Bob use the millionaire protocol to compare s_A−r_A to r_B−s_B. This is equivalent to comparing the squared Euclidean distance d(x, y_i) to r. If s_A−r_A<r_B−s_B, then Bob return u_i+1, otherwise Bob returns u_i. Alice stores the result in v_i.
    • (b) Alice and Bob determine

$v=\sum _{i=1}^N{v}_i$

• • •  and

$u=\sum _{i=1}^N{u}_i,$

• • •  respectively.
    • (c) Alice and Bob use the millionaire protocol to compare v to u+k and do the following:
      • i. If v>u+k, then r is too large, and either Alice decreases r_A by 1 or Bob decreases r_B by 1, and repeat.
      • ii. If v<u+k, then r is too small, and either Alice increases r_A by 1, or Bob increases r_B by 1, and repeat.
      • iii. If v=u+k, terminate.

EFFECT OF THE INVENTION

The invention provides a method for classifying private information using various dot product based classifiers, such as support vector machines (SVM), neural-networks, AdaBoost, and k-nn classifiers.

It is to be understood that various other adaptations and modifications may be made within the spirit and scope of the invention. Therefore, it is the object of the appended claims to cover all such variations and modifications as come within the true spirit and scope of the invention.

Claims (6)

1. A method for securely classifying private data of a first party using a classifier H(x) of a second party, comprising a client and a server for performing steps of the method, comprising the steps of:

(1) providing, using the client, private data of a first party as a vector x, and providing, using the server, a classifier

$H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x\right)\right)$

by a second party, where

$h_n\left(x\right)=\{\begin{array}{cc}{\alpha }_n& x^T{y}_n>{\Theta }_n\\ {\beta }_n& \mathrm{otherwise}\end{array};{\alpha }_n,{\beta }_n$

and Θ_n are

scalar values; and y_n is a vector storing parameters of the classifier, and generating, by the second party, a set of N random numbers, s₁, . . . , s_N, such that

$s=\sum _{n=1}^N{s}_n;$

(2) performing, for each n=1, . . . , N, the sub-steps:

(a) applying a secure dot product to x^Ty_n to obtain a_n for the first party and b_n for the second party;

(b) applying a secure millionaire protocol to determine whether a_n is larger than Θ_n−b_n, and returning a result of a_n+s_n or β_n+s_n;

(c) accumulating, by the first party, the result in c_n; and

(3) applying the secure millionaire protocol to determine whether

$c=\sum _{n=1}^N{c}_n$

is larger than

$s=\sum _{n=1}^N{s}_n,$

and returning a positive sign if true, and a negative sign if false to classify the private data x, wherein the providing, the performing, and the applying steps are performed in part by the client and in part by the server using secure multi-party protocols.

2. The method of claim 1, in which H(x) is a strong classifier that is a linear combination of weak classifiers h(x).

3. A method for securely classifying private data of a first party using a classifier H(x) of a second party, comprising a client and a server for performing steps of the method, comprising the steps of:

(1) providing private data of a first party as a vector x, and providing a classifier

$H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x\right)\right)$

by a second party, where h_n(x^Ty_n)=(x^Ty_n+c)^d, where c and d are scalar values, and y_n is a vector storing parameters of the classifier, and generating, by the second party, a set of N random numbers, s₁, . . . , s_N, such that

$s=\sum _{n=1}^N{s}_n;$

(2) performing, for each n=1, . . . , N, the sub-steps:

(a) applying a secure dot product to x^Ty_n to obtain a_n for the first party and b_n for the second party;

(b) constructing, by the second party, a polynomial

ƒ_n(a)=(a+b+c)^d+s_n=(x^T y_n+c)^d+s_n;

(c) applying an oblivious polynomial evaluation to ƒ_n(a);

(3) determining, by the first party,

$r=\sum _{n=1}^N{f}_n;$

(4) applying the secure millionaire protocol to determine whether r is larger than s, and returning a positive sign if true, and a negative sign if false to classify the private data x, wherein the providing, the performing, the determining and the applying steps are performed in part by the client and in part by the server using secure multi-party protocols.

4. A method for securely classifying private data of a first party using a classifier H(x) of a second party, comprising a client and a server for performing steps of the method, comprising the steps of:

(1) providing private data of a first party as a vector x, and providing a classifier

$H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x\right)\right)$

by a second party, where h_n(x^Ty)=exp(γ∥x−y_n∥₂ ²), and y is a vector storing parameters of the classifier and generating, by the second party, a set of N random numbers, s₁, . . . , s_N, such that

$s=\sum _{n=1}^N{s}_n;$

(2) performing, for each n=1, . . . , N, the sub-steps:

(a) selecting, by the second party, a random number s_B and constructing a function ƒ(z)=γz−s_B, where z is a scalar.

(b) applying oblivious polynomial evaluation to obtain a share s_A for the first party, where s_A=ƒ(x^Tx) is a secure dot product;

(c) applying the secure dot product to x^T(−2γy_n) to obtain a share r_A for the first party and a share r_B for the second party;

(d) constructing, by the first party, t_A=exp(s_A)exp(r_A);

(e) constructing, by the second party, t_B=exp(s_B)exp(r_B)exp(γy_n ^Ty_n);

(f) applying the secure dot product to t_At_B=exp(s_A+r_A+s_B+r_B+γy_n ^Ty_n)=exp(γ∥x−y_n∥₂ ²) to obtain a_n, for the first party and b_n for the second party;

(3) determining an accumulation

$a=\sum _{n=1}^N{a}_n$

for the first party, and an accumulation

$b=\sum _{n=1}^N{b}_n$

for the second party;

(4) applying the secure millionaire protocol to determine whether a is larger than b, and returning a positive sign if true, and a negative sign if false to classify the private data x, wherein the providing, the performing, the determining and the applying steps are performed in part by the client and in part by the server using secure multi-party protocols.

5. A method for securely classifying private data of a first party using a classifier H(x) of a second party, comprising a client and a server for performing steps of the method, comprising the steps of:

(1) providing private data of a first party as a vector x, and providing a classifier

$H\left(x\right)=\mathrm{sign}\left(\sum _{n=1}^N{h}_n\left(x\right)\right)$

by a second party, where h_n(x^Ty_n)=1/(1+exp(x^Ty_n)), and y_n is a vector storing parameters of the classifier;

(2) performing, for each n=1, . . . , N, the sub-steps:

(a) applying a secure dot product to x^Ty_n to obtain private shares a_n for the first party and b_n for the second party;

(b) applying the secure dot product to s_A+s_B=exp(a)exp(b) to obtain s_A for the first party and s_B for the second party;

(c) selecting, by the second party, a random number r_B, and defining ƒ(s_A)=r_B+r_B(s_A+s_B)=r_B(1+exp(x^Ty_n));

(d) applying an oblivious polynomial evaluation to obtain r_A=ƒ(s_A) for the first party;

(e) applying the secure dot product to

$\frac{1}{r_A}{r}_B=\frac{1}{r_B\left(1+\exp \left(x^T{y}_n\right)\right)}$

to obtain t_A,n for the first party and t_B,n for the second party;

(3) determining

$t_A=\sum _{n=1}^N{t}_{A,n}$

by the first party, and

$t_B=\sum _{n=1}^N{t}_{B,n}$

by the second party; and

(4) applying the secure millionaire protocol to determine whether t_A is larger than −t_B, and returning a positive sign if true, and a negative sign if false to classify the private data x, wherein the providing, the performing, the determining and the applying steps are performed in part by the client and in part by the server using secure multi-party protocols.

6. A method for securely classifying private data of a first party, comprising a client and a server for performing steps of the method, comprising the steps of:

(1) providing a private data vector x, a private share r_A of a radius r of a first party, and points y_i and a private share r_B of the radius r of a second party;

(2) performing by the first party and the second party for each point y_i the sub steps:

(a) obtaining by the first party and the second party private shares a and b, respectively, of the dot product (−2x)^Ty;

(b) defining by the second party s_B=b+y^Ty a function ƒ(z)=z+s_B, where z is a scalar;

(c) applying an oblivious polynomial evaluation to evaluate s_A=ƒ(x^Tx+a), where s_A is a private share of the first party;

(d) selecting by the second party a random number u_i;

(e) comparing s_A−r_A to r_B−s_B using a millionaire protocol, and if s_A−r_A<r_B−s_B AND c(y_i)=1, returning u_i+1, and otherwise the second party returns u_i. The first party stores the result in v_i;

(3) determining

$v=\sum _{i=1}^N{v}_i$

and

$u=\sum _{i=1}^N{u}_i$

by the first party and the second party, respectively;

(4) comparing v and u+k/2 using the millionaire protocol, and if v>u+k/2, then x is labeled as 1, otherwise as 0, wherein the providing, the performing, the determining and the comparing steps are performed in part by the client and in part by the server using secure multi-party protocols.

Images